A month before the Brexit vote, the Ipsos Mori polling firm discovered that the British public was systematically misinformed. For instance, Leave voters believed that EU immigrants comprise 20 per cent of the UK’s population. Remain voters estimated 10 per cent. The truth is about 5 per cent. Both Leave and Remain voters vastly overestimated what percent of the UK child benefit goes to children living in Europe. Both vastly underestimated how much foreign investment comes from the EU, and overestimated how much comes from China.
Brexit voters were misinformed about the basic facts, though Remainers on average were less wrong than Leavers. Yet even knowing these basic facts would not suffice to make voters well informed. To cast a smart vote, a citizen would need significant social scientific knowledge. They would need to know about the economics and sociology of trade and immigration, the politics of centralised regulation, and the history of nationalist movements. But there is no reason to presume even a tenth of the UK’s population has a basic grasp of the social science needed to evaluate Brexit.
Brexit illustrates the central problem with democracy. Political decisions are high stakes. To make sound decisions requires knowledge. But voters nearly always lack the necessary knowledge. They are not just ignorant; they are usually misinformed. Democracy is the will of the people, but the people usually have little clue what they are doing.
The fact of political ignorance
In designing political institutions, we face an unhappy dilemma. If we concentrate power in the hands of the few, their individual inputs make a significant difference. They thus have every incentive to use their power in a rational and informed way, but also to use their power selfishly. On the other hand, if we spread power out widely among the many, each individual loses the incentive to act selfishly, but also loses the incentive to be informed or to think in a rational, scientific way. Representative parliamentary government is supposed to split the difference, by making the elites answer to the people while also empowering them to override the people’s choices. Of course, referendums attempt to bypass this check on voter irrationality.
Thousands of years ago, Plato hypothesised that the democratic masses would be too ignorant to run their government. He did not have the data to back up his assertion. We do. It turns out Plato was right.
For over 65 years, researchers in the United States have been collecting data about what citizens know and don’t know. The results are depressing. (Results for other countries are similarly depressing; I use American examples because we have more data.) The American National Election Studies, conducted every other year, gives voters a test of basic political information. In general, the top 25 per cent of voters get about 90 per cent of the questions right, the next 50 per cent do little better than chance, and the bottom 25 per cent do worse than chance. In a typical election year, most Americans cannot identify their congresspeople or which party controls congress. Most cannot guess the unemployment rate within a few percentage points. Most have little memory of recent events.
In the 2000 US presidential election, while most Americans knew that Al Gore was more “liberal” than Bush, most did not seem to know what that term means. Fewer than half understood that Gore was more supportive of abortion rights, more supportive of welfare-state programmes, favoured a higher degree of aid to black people or was more supportive of environmental regulation.
Voters do not know the basic facts, but even if they did, that would not be enough. People choose candidates or parties who push policy agendas. Even if, contrary to fact, voters knew what candidates or parties wanted to do, they would still need to understand some basic social science in order to determine which policies are most likely to deliver their favoured ends. In the same way, even if I know I have stomach ulcers, I do not magically acquire the knowledge needed to treat the ulcers.
In fact, there is a big disparity between how laypeople and economists (both left and right) think about the economy. This disparity is not explained by demographic differences. Further, it turns out that better informed or higher IQ laypeople tend think more like economists.
It is no mystery why democratic voters know so little. The are not stupid; they just do not care. The democratic process gives them bad incentives.
In a democracy, how the majority as a collective votes matters, but how any individual votes does not. An individual vote matters only if it decides the election or the referendum, that is, only if it breaks a tie. Yet the probability that a vote will break a tie in a major election or referendum is vanishingly small. Accordingly, each individual voter has little incentive to be informed. They can afford to be ignorant, to be misinformed, or to indulge grossly irrational beliefs. For most of them, the cost of being informed outweighs the potential benefits.
As an analogy, suppose a university lecturer informed a 47 million-person class that they would take a final exam in three months, an exam worth 100 per cent of their final grade. But suppose she told them, “You will not receive your own grade. Instead, I shall average your grades together and you will each receive the class average.” In such a scenario, the class would likely fail. That is how democracy works, except that we cannot even force people to listen to three months of expert lectures before they take the test.
To be clear, it is not as though whatever the majority of the electorate prefers on election day automatically gets translated into policy. Even the Brexit referendum was not binding, the High Court has ruled parliament must take a vote and the Supreme Court could follow suit. More broadly, members of parliament, ministers, and bureaucrats have significant independence. Nevertheless, politicians tend to give voters what they want. But they want what they want because they do not know what they need to know.
The foolishness of crowds
Some defenders of democracy downplay the problems of democratic ignorance. They claim that wisdom is an emergent feature of crowds. Sure, they say, individuals may be ignorant (or worse), but when we aggregate their ignorant beliefs together, we get good results.
This theory sounds delightful, but it holds only in special cases.
Here are some such cases. If we ask carnival goers to estimate the number of jelly beans in a jar, most estimates will be far off. However, the more people we ask, the more accurate the average estimate becomes. Similar results hold for guessing the height of the prize horse or the weight of the prize pig.
If only citizens on the whole were as good at choosing members of parliament as they are at guessing the weight of prize pigs. The problem is that in politics, citizens make systematic mistakes, whereas in guessing the weight of pigs, they make random mistakes which cancel each other out. The so-called “wisdom of crowds” appears only in situations where individual errors are largely random.
In politics, alas, errors are not random. For instance, in the Brexit vote, Leave voters vastly overestimated the number of EU immigrants, while Remain voters slightly overestimated the number.
It is comforting to believe that perhaps the Labour and Conservative Parties will cancel each other’s mistakes out. Yet there is little reason to think the truth lies somewhere in-between. The spectrum between Labour and Conservative political beliefs occupies only a narrow niche in ideological space.
The same point holds for Democrats and Republicans in the United States. In the recent US presidential election, Trump pushed a strongly anti-immigrant agenda, while Clinton pushed to keep the already highly restrictive status quo. When we ask economists, though, we find that both major candidates and parties are far more restrictionist than the evidence supports.
Do not blame the schools
Commentators often think there is an easy solution to the problem of democratic ignorance: education. This just shows we need more and better public education, they say. We need to teach schoolchildren critical thinking, philosophy, politics, and economics.
It is a natural response, but it misses the point. The schools are not to blame. The problem is incentives.
In secondary school, citizens already know most of what they need to know to be good voters. They promptly forget it after their exams end. In fact, studies generally find that most citizens retain only a small fraction of the information they learned in schools.
Citizens allow themselves to forget the information because they do not find it interesting or useful. This type of knowledge is not useful because their individual votes count for so little. If I forget how to drive a car, I cannot get to work. But if I forget the economics of trade and vote accordingly, well, nothing happens.
British citizens are more educated than ever before. But in the UK, as elsewhere, levels of political knowledge have remained stable, and low, over the past half century. Even the internet has made little difference. As the joke goes, we carry devices with us capable of accessing all the world’s knowledge in an instant, but we use them to share pictures of our food.
The injustice of democratic incompetence
Democracies are not just about choosing mundane things like flag colors and national anthems. Democratic decisions affect matters of life and death, peace and war, prosperity and poverty.
The damage of bad decisions can compound over time. Suppose Brexit turns out to reduce the UK’s economic growth by one percentage point year after year. A one percentage point reduction in growth in any given year might not seem trivial. But over time, trivial problems add up to big problems. Had the UK had one percentage point less average growth since 1948, its gross domestic product would be only about £600bn, and the standard of living would be as low as Egypt’s.
Incompetent decision-making is not merely bad; it may be unjust. Consider an analogy. Suppose a 12-person jury hears a murder trial. Suppose the jury is ignorant: they completely ignore the facts of the case entirely, but decide to find the defendant guilty on the basis of a coin flip. Suppose instead they are irrational: they hear the facts of the case, but decide to find the defendant guilty because they believe him secretly to be an evil robot. Suppose instead they are malevolent: they believe the evidence indicates the defendant is innocent, but find him guilty because they hate people of Indian heritage. Or, finally, suppose instead the jury acts in bad faith: They find the defendant guilty because they accepted a bribe.
If we knew a jury behaved in any of these ways, we would judge the trial unfair and unjust. The defendant’s freedom and welfare is at stake. They owe the defendant, and perhaps the British public whom they represent, to decide the case competently and in good faith.
This line of reasoning applies just as well to the electorate. Political decisions are also high stakes. The outcomes — including all ensuing laws, regulations, taxes, budget expenditures, wars, and so on — are imposed upon us involuntarily. These decisions can and do harm us, and can and do deprive many of us of property, liberty and even life. At first glance, we should think that voters, like jurors, have a moral obligation to vote in a competent and morally reasonable way. But it appears that voters systematically violate this obligation.
Some might say this analogy fails. They claim voters are only hurting themselves. But that is not true. If I eat a dozen Cadbury Dairy Milks per day, I may get diabetes, but only I suffer the consequences. When voters vote, the majority decides not just for itself, but for the dissenting minority, and for the tens of millions affected by the decision who cannot vote. The disproportionately elderly Leave voters wish to impose their will upon the disproportionately young Remain voters.
We cannot fix democracy
The problems of political ignorance, misinformation, and irrationality are built-in to democracy. They are not superficial flaws, but problems democracy itself creates. We cannot “fix” the problem with more democracy anymore than we can fix a flood with more rain.
At the very least, these problems highlight why representative parliaments are important and why referenda are, in general, terrible ideas (especially in large democracies). We want our representatives to represent us. But sometimes that does not simply mean promoting the policies we favour. Sometimes, instead, it means doing what is good for us rather than what we want. Members of parliament are much better informed than the public as a whole, and they recognise their decisions have real weight.
A recent study on US politics seems to confirm that representative government explains why democracy performs as well as it does. Princeton University political scientist Martin Gilens asked, when voters at the 10th, 50th, and 90th percentiles disagree about policy, with whom do US presidents side? He found that presidents were about six times more likely to implement policies high-income voters want than what middle- or low-income voters want. (Surprisingly, George W Bush was more likely to side with low-income voters than more left-wing presidents.) We might say that this shows democracy does not work. But there is an upside: As Gilens notes, high-income voters are also in general high-information voters, and low-income voters are also in general low-information voters. Even within the Democratic Party, high- and low-information voters have different policy preferences. High-information voters are far more in favour of free trade and immigration, less in favour of war, and far more protective of civil liberties. Perhaps representative democracies overperform because politicians do not quite have to answer to the median voter.
If we want even better performance, though, we might need to depart from democracy altogether. The economist Robin Hanson argues we should use specialised betting markets to choose policies. We should, he says, bet on beliefs but vote on values. Law professor Ilya Somin of George Mason University argues that we need a system in which people can “vote with their feet”. This means having radically devolved political systems so that citizens can choose where to live. Both Hanson’s and Somin’s ideas mean to overcome voter ignorance by making mistakes costly and truths beneficial. University of Southampton philosopher Ben Saunders argues we should bypass voting and instead choose leaders through sortition, the random selection system used by the Athenians. Finally, in my recent book Against Democracy, I argue (expanding on an idea from John Stuart Mill) that we should consider implementing electoral systems in which citizens’ votes are weighted by their degree of objective political knowledge.
These proposals set off alarm bells, especially among people who have not studied them in any depth. But they at least illustrate our dilemma. A misinformed electorate is the price we pay for universal, equal suffrage.
(Jason Brennan is a professor of strategy, economics, ethics, and public policy at Georgetown University, Washington DC.)